Modelpublic

Virros/Vistral_Function_Calling_500

sourceHugging Faceafl-3.0updated 2y agoView on Hugging Face

0likes71downloads

Model Card

VistralFunctionCalling_500

This model is a fine-tuned version of Viet-Mistral/Vistral-7B-Chat on the generator dataset. It achieves the following results on the evaluation set:

—Loss: 0.2455
—Rouge1: 0.8798
—Rouge2: 0.7704
—Rougel: 0.8144
—Rougelsum: 0.873
—Gen Len: 2048.0

It achieves the following results on the test set:

—Loss: 0.2639
—Rouge1: 0.8874
—Rouge2: 0.7745
—Rougel: 0.8141
—Rougelsum: 0.8811
—Gen Len: 2048.0

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

—learning_rate: 0.0002
—trainbatchsize: 3
—evalbatchsize: 2
—seed: 42
—gradientaccumulationsteps: 2
—totaltrainbatch_size: 6
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: cosine
—lrschedulerwarmup_ratio: 0.03
—num_epochs: 2
—mixedprecisiontraining: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
0.8125	0.25	3	0.8199	0.8499	0.7054	0.7623	0.8429	2048.0
0.658	0.5	6	0.3952	0.8634	0.7361	0.7854	0.8564	2048.0
0.4082	0.75	9	0.3261	0.8732	0.7452	0.7927	0.8657	2048.0
0.3302	1.0	12	0.2928	0.8733	0.7552	0.801	0.8666	2048.0
0.2653	1.25	15	0.2653	0.8775	0.7646	0.809	0.8703	2048.0
0.2605	1.5	18	0.2528	0.8778	0.7678	0.8119	0.8707	2048.0
0.2444	1.75	21	0.2476	0.8793	0.7697	0.8132	0.872	2048.0
0.23	2.0	24	0.2455	0.8798	0.7704	0.8144	0.873	2048.0

Framework versions

—PEFT 0.11.1
—Transformers 4.41.1
—Pytorch 2.3.0+cu121
—Datasets 2.19.2
—Tokenizers 0.19.1